[Gluster-users] georeplication on huge files

Venky Shankar vshankar at redhat.com
Wed Jul 18 04:05:07 UTC 2012


On 07/18/2012 03:14 AM, Gandalf Corvotempesta wrote:
> how does gluster perform on huge file georeplication like a VM disk
> based image?
>
> Let's assume a 150GB image on a single file. Does it replicate the whole
> file every time, like rsync, or only differences will be replicated?

AFAIK, Rsync does not blindly sync whole file. It uses rolling checksum 
algorithm to decide which blocks to sync.

Geo replication relies on Rsync to sync file contents after it has 
identified that need synchronization.

>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>


Thanks,
--Venky




More information about the Gluster-users mailing list